The economic system of record for the lab

Most labs run on roll-up P&L and gross margin. Not because anyone wants to — because nothing shows them more.

Your lab’s economics live in the gaps between your systems. BenchForge models your whole lab as one path — every step, instrument, handoff and integration — costs each node, finds what’s actually limiting you, and grades every decision against what happened. It runs on the exports your LIMS and billing systems already produce, keeps PHI out by design, and replaces nothing.

When the 2027 rates post this fall

Load the published file and see your whole menu re-priced against it — which lines move, which way, and which stop covering their cost. Not a forecast of what CMS will do: your own menu, run against the actual posted numbers, in one working session. You can try that on the sample lab right now.

Live demo · free · no signup
Open a fully-modelled sample lab
A real lab with ten assays, its own P&L and a decision on the record — not a video, not a sandbox. Pick where you want to start and we walk you through it. Modelling your own lab gets you a first costed answer in minutes.
Already know what you came for?
Your discipline: Clinical genomics / molecular · Hereditary / germline · Broad-menu / regional · Research / core facility
Or open a sample lab directly: · · ·
  • ✓Cost per test you can defend
  • ✓The binding constraint, costed
  • ✓Plan vs. actual over time

Said plainly: the first two you can watch work right now on a fully-modelled sample lab. The third takes a quarter — a decision has to be committed and the actuals have to arrive before anything can be graded, so until a design partner completes a cycle, what we can show you there is a seeded example.

What you’ll see in 10 minutes
Cost / test · CGPtumor + matched normal — the whole report$1,355
Net marginlab-wide, all assays25%
Binding constraintPinpointed & costed
Automation payback14 mo

Figures from the Clinical NGS demo below — open it and check them. Modeled estimates, not financial advice; validate independently.

See it in action
Worked example · automation decision

“Should we buy an automated liquid handler to take library prep off the bench?” BenchForge models the manual prep step today — touch-time, FTE load, error/rework, the throughput it caps — then the same workflow with the handler in place, against the capex. You get recovered $/month, FTE freed, the new bottleneck, and payback / NPV / IRR — a board-ready business case instead of a gut call.

Pricing
Concierge alpha — two founding partners comped, everyone else pays
When it opens: $1,000/month for year one, then $2,000/month.
Unlimited users. Advisor and consultant seats free. No per-assay metering.
See what’s included →
Your data stays yours

Each workspace is isolated by row-level security — no other customer can see your models, and we don’t use your numbers or models to train anything or sell them. BenchForge models economics, not patient data: it is not designed for PHI, and inputs are aggregates (volumes, costs, reimbursement, turnaround). Keeping PHI out is your responsibility — automated screens back you up on data import: uploads are checked for PHI-named columns and identifier-shaped values; the API and AI-parse paths reject on a hit (nothing is persisted), and the in-browser CSV import flags what it finds for you to remove. To improve the product we collect anonymous usage analytics (which features are used, where people get stuck) — never your numbers, models, or identity. Do-Not-Track is honored.

BenchForge is a decision-support and economics-modeling tool — not a clinical, diagnostic, or LIS system.

Getting started

What you’ll need to validate your model

The complete input set. Start from a template, swap in your numbers (required + recommended), then reconcile one month of actuals to validate. Required inputs gate whether the economics can be trusted.

1

Start from a template

~2 min · no data needed

Pick the workflow you run (WGS, WES, targeted panel, RNA-seq, RT-qPCR, PGx) — or start blank. Every step, instrument, cost, and turnaround comes pre-filled from cited public sources, so you get a first answer before entering anything.

2

Make it your lab

~10–15 min · per workflow

Replace the template defaults with your numbers. These four are required before the economics can be trusted; the rest sharpen the picture.

  • requiredMonthly volume— samples/month for each workflow you run
  • requiredReimbursement— a flat $/test, or a payer mix with each payer’s collection rate
  • requiredSpecimen type— FFPE, blood, swab, cfDNA… — drives expected re-extraction/QC loss
  • requiredLibrary-prep / key reagent cost— $/sample for your main prep kit
  • recommendedSequencing depth— coverage × genome size, or Gb/sample
  • recommendedLoaded labor rate— fully-burdened $/hour for your techs
  • recommendedTurnaround SLA target— the TAT you commit to per workflow — lights up SLA health
  • recommendedStaffing & operating schedule— FTEs/hours per role and your operating days/hours (lab-level)
  • optionalCompliance cost— proficiency testing + competency, $/month (lab-level)
3

Reconcile against your actuals

~5 min · the trust step

This is what turns a reviewed model into one reconciled against reality. Import one month of real numbers and BenchForge reports "modeled within X% of actual" — a credibility check for your leadership team.

  • requiredActual monthly COGS— total cost of goods for the month ($)
  • requiredActual monthly tests— count of tests reported out that month
  • requiredMonth— which month the actuals are from

Reviewed vs reconciled: entering your numbers (Stage 2) makes the model yours; reconciling actuals (Stage 3) shows how closely it tracks reality. BenchForge is an informational modeling tool — validate independently before any capital, staffing, or operational decision.